3.1 Overview of the established incumbents in the global x-ray scanners market

The products at the heart of the dispute are the x-ray machines for scanning large cargo items such as railroad cars and trucks. These scanners use relatively sophisticated technology and are purchased primarily by governments for border security measures. Depending on the size and use of the machines, the technology employed differs. The machines for large objects use high-energy beams; similar scanners for hand baggage utilize lower-energy x-rays. This difference will prove to be important later for the case.

In the wake of the 11 September 2001 terrorist attacks, demand for x-ray scanners increased. Not surprisingly, as the market grew, international competition heightened. Until the mid-2000s, however, the global market for x-ray scanners was dominated primarily by a handful of producers in the developed world. In recent years, a Chinese producer has managed to gain share in this growing market, at least partly as a consequence of Chinese government industrial policy. Indeed, the product's technical sophistication, its high start-up costs, and the oligopolistic nature of the industry fit well into the scenario envisioned in many theoretical setups for strategic industrial policy.

The story here is of a battle between two rivals – a European incumbent and a Chinese upstart – for the global x-ray scanner market. Antidumping duties surface as part of the struggle between the two, with each appealing to its government to take action to keep the other out of its home market.

The incumbent is Smiths Heimann GmbH (‘Smiths’), the German branch of Smiths Detection PLC, an EU-based multinational company housed within the Smiths Group PLC, a European conglomerate. By the late 2000s, Smiths held a dominant position within the European market, accounting for approximately 80% of total EU production of the relevant x-ray scanners.Footnote ¹⁴ How was Smiths able to achieve such a dominant hold over the European market? Part of the story may very well be the high quality of its products. But Smiths' own annual reports hint at another possible factor that may have played a role. As noted in several of its annual reports, the vast majority of Smiths' sales are to governments and government agencies around the world. Some European governments may very well have preferred to purchase from a European producer. The fact that Smiths resides within a pan-European conglomerate, coupled with its long-time presence in the sector, likely made Smiths an attractive recipient of European government procurement contracts for x-ray scanners.Footnote ¹⁵

Smiths is also a market leader worldwide, but its global share is much less dominant than in Europe. In 2008, Smiths led the world with a 29% market share in scanners for the ‘explosives, weapons, and contraband equipment’ sector.Footnote ¹⁶ The global market is much more fragmented for x-ray scanners for ports and borders (the relevant sector for this case). Smiths' traditional competitors in this sector are primarily US-based multinationals.Footnote ¹⁷ Today, according to Smiths, they include Rapiscan, American Science and Engineering (ASE), and Leidos (formerly ‘SAIC’).Footnote ¹⁸ Smiths and this group of American companies dominated sales in the growing market for x-ray scanners until the mid-2000s.

3.2 China's quest to build a latecomer entrant through strategic industrial policy

The emergence of a Chinese producer in the past decade has resulted in a major disruption to the landscape of players competing for a share of the global market for x-ray scanners. In many ways, this story is similar to the narrative that has played out for many other industrial products. However, in this instance, the disruption began with domestic sales rather than through an export-led strategy.

Over the past two decades, China has invested vast sums to upgrade its infrastructure. These investments, in turn, have generated tremendous demand for x-ray scanners to safeguard Chinese ports, roads, airports, railway hubs, subways, etc.

In some sectors, China has willingly turned to foreign imports to meet its growing demand. In others, however, China has attempted to create a national champion.Footnote ¹⁹ Because of the underlying importance of security to the regime, it would not be surprising for the Chinese government to view the x-ray scanners market as one in which they would prefer a domestic ‘champion’. We are not aware of any available government document that explicitly sets forth such a strategy but several interesting facts point strongly to this possibility.

Segal (Reference Segal2010) notes that Beijing's strategy has been to rely on a mixture of top-down and bottom-up initiatives focused on developing capacity in ‘fields that affect the lifeblood of the national economy and national security’. Certainly, security scanners for infrastructure would appear to qualify as such a field. Bottom-up initiatives have centered on university–industry collaboration and venture capital. That precisely has been the approach used to cultivate a domestic Chinese producer to challenge the established European and American players in the x-ray scanners market.

The Chinese firm that has become a major global player is Nuctech Company Ltd. (‘Nuctech’). It is, however, not just another Chinese company. Nuctech was created in 1997 at Tsinghua University, China's premier university for science and technology.Footnote ²⁰ Even after the firm was spun off, Tsinghua University remains active in it; it is the leading shareholder in Nuctech through its investment holding company, Tsinghua Holdings. As in many other start-ups, Nuctech's CEO during the course of its initial rapid growth was a dynamic young man named Hu Haifeng. Hu is a Tsinghua alumnus, having received his executive MBA there. He also is the son of Hu Jintao, China's top leader from 2002 to 2012 and prior to that, its Vice President. In 2008, based on his success at Nuctech, Hu Haifeng was promoted to become the Communist Party secretary of Tsinghua Holdings.

Nuctech grew by executing a strategy modelled on the one employed by Smiths in Europe. It sought to become the dominant producer within its home market and then leverage this dominance to become a major player worldwide. This strategy proved effective at least in part because of the dominance of government procurement contracts in this market and because of Nuctech's close ties with the government.

Just as European governments may prefer to buy x-ray scanners from a European company, Nuctech counted on the fact that the Chinese government too would prefer to buy from a Chinese company, especially one run by the President's son. Furthermore, as China is not a party to the GPA, its government's actions in this sector are even less constrained compared to its European and American competitors (whose governments have signed on to the GPA). In the late 2000s, Nuctech accounted for approximately a 90% share of the domestic Chinese market (Erickson 2010).Footnote ²¹

Like Smiths, Nuctech took advantage of its strong position in its home market to expand overseas. In 2001, Nuctech won a bid to install x-ray scanners at the ports of Sydney and Melbourne. By 2008, Nuctech had won contracts in over 60 countries, many in the developing world (Goodwin Reference Goodwin2008). This included supplying x-ray scanners for the port of Los Angeles, a contract that Nuctech won by outbidding Smiths and SAIC.

What accounts for Nuctech's success? Part of it may be due to Nuctech's ability to develop a lower-cost alternative to the x-ray scanners offered by incumbents. For several contracts, Nuctech outbid its competitors on the basis of price. But industrial policy also likely played a role. Government agencies account for the vast majority of China's purchases of x-ray equipment, and those agencies delivered business to Nuctech. Hints of an industrial policy agenda also can be found in a Tsinghua University press release trumpeting Nuctech's early success in Australia. It describes the deal as ‘a successful step in the Chinese government's strategy to translate technological advantages into commercial uses’.Footnote ²²

Nuctech, however, has sought to distance itself from the Chinese government. In explaining why Nuctech was awarded a contract to supply scanners to detect liquid explosives to all 147 airports in China ahead of the 2008 Olympics, an official noted that Nuctech's products were deemed the most satisfactory. While acknowledging that Nuctech did have connections to the President's son, the official said that this ‘had no effect on our decision’, adding that ‘[t]his has to do with people's lives, and you can't play games with safety’.Footnote ²³ Similarly, authorities for the port of Los Angeles, in the wake of their controversial decision to use government funds to buy Chinese products from President Hu's son's company, also explained that their decision was driven primarily by price.Footnote ²⁴

Nuctech's competitors, however, contend that ties between the company and the Chinese government remain strong. They allege that the Chinese government provides soft loans to allow Nuctech's customers to pay for the x-ray scanners and that government policies allow Nuctech to drive down price to win contracts (Aldrick et al., Reference Aldrick2009). However, despite these allegations, no government to date has challenged China on this matter before the WTO.Footnote ²⁵

Nuctech's rise nicely illustrates two points: First, the story is consistent with theories of strategic trade policy. The Chinese government nurtured an ‘infant industry’ in a sector with increasing economies of scale through import protection. The government's industrial policy allowed Nuctech to increase production and lower its costs by keeping foreign products largely out of the growing Chinese market. This, in turn, boosted Nuctech's export competitiveness and allowed Nuctech to emerge as a lower-cost competitor to the established incumbents.

Second, antidumping duties do not feature as part of the Chinese government's initial industrial policy. Import protection is achieved through other trade-related policy instruments besides antidumping – namely, government procurement and indirect export subsidies.Footnote ²⁶ To the extent that the Chinese government played a role in nurturing Nuctech's success, it was not through the use of antidumping duties. The conventional account of how antidumping duties intersect with strategic industrial policy therefore is not what played out in this instance.

3.3 EU Antidumping action to ward off Chinese entrant

How then did antidumping duties enter into the picture? As suggested in section 2, an alternative mechanism is possible – one which is firm-led in reaction to an industrial policy. This is precisely what happened in this instance. By the 2000s, Nuctech had started to make impressive inroads worldwide, including against Smiths in its European home market. For example, in 2006, the port of Dublin awarded Nuctech a contract to supply x-ray scanners for its facilities. Nuctech's share of the x-ray scanner market in Europe more than doubled between 2004 and 2006.Footnote ²⁷ A critical component of Nuctech's successful strategy was its price, which reportedly was 15 to 20% lower than that for the comparable Smiths product.Footnote ²⁸ Smiths decided that it was time to take legal action.

As discussed earlier, antidumping duties provide a means for an incumbent to seek help from its government in warding off a foreign upstart. The incumbent accuses the foreign entrant of ‘dumping’ its product by selling it below the ‘normal value’ of the product in its home market. This strategy is relatively easier to employ when the foreign entrant is Chinese, because China's Protocol of Accession allows its trading partners to classify China as a ‘non-market economy’ (NME) when undertaking an antidumping investigation.Footnote ²⁹ This NME status often results in dumping margins much higher than those normally calculated by the investigating agency. This in turn can lead to much higher duties if material injury is found to have been caused by the dumped Chinese imports.

Smiths brought an antidumping case against ‘Certain Cargo Scanning Systems’ from China on 2 February 2009. Note that while Smiths' petition called for an investigation of all such products from China, the single relevant exporter was Nuctech. On 18 March 2009, the European Commission (‘Commission’) initiated an investigation into Nuctech's pricing of all cargo-scanning systems with x-rays of 250 KeV (kiloelectronvolts) or higher.Footnote ³⁰ The Commission established that the period of investigation (POI) would be from 1 July 2007 to 31 December 2008.

The Commission announced a preliminary 36.6% antidumping duty against Nuctech on 16 December 2009. The final duty was 34% and came into force on 14 June 2010. With the imposition of the antidumping duties, Nuctech's pricing advantage over Smiths was reduced significantly. Smiths had managed successfully to use the European antidumping system to raise costs to a new Chinese rival as a concerted strategy to preserve its dominant position within the EU market.

Note that the emergence of antidumping measures in this instance was not part of a government-led strategic industrial policy. Instead, it was part of a firm-led initiative by the incumbent to ward off an entrant supported by industrial policy. The Commission, for its part, responded because it was forced to do so after receiving Smiths' petition and not necessarily as part of an active strategy to provide support to an embattled domestic producer.

3.4 Chinese antidumping action in response

After being hit with antidumping duties in an export market, note that Nuctech had at least two potential responses (which are not mutually exclusive). One option would be to lobby the Chinese government to file a WTO dispute against the government imposing the duties. A second option, discussed earlier, is to file its own antidumping case in its home market against its competitor in order to trigger tit-for-tat retaliatory duties. For now, we refrain from a discussion of other potential options.Footnote ³¹

Had Nuctech pursued the first option, it would have lobbied the Chinese Ministry of Commerce (‘MOFCOM’) to file a WTO dispute against the EU for the imposition of its antidumping duties. In recent years, China has filed several such cases and prevailed on multiple occasions.Footnote ³² However, this strategy is subject to several constraints and potential disadvantages. The most important of these is that the government must be able to craft a plausible argument that the Commission's actions were inconsistent with the EU's WTO obligations. If not, then no matter how hard Nuctech lobbies, MOFCOM might not be willing to expend resources for a case. Secondly, WTO disputes typically take multiple years to resolve. Firms, in the meantime, must be able to endure the impact of the higher antidumping duties, while their competitors benefit from the relative change in costs.

The second option, which can be executed as an alternative or in addition to the first option of a WTO dispute, was for Nuctech to initiate its own petition with the Chinese government to impose antidumping duties against Smiths in its own domestic market. This strategy has several advantages. First, the time necessary for a domestically administered antidumping case is far shorter than for a WTO dispute. Second, preliminary duties can be imposed before the conclusion of the case, meaning that the case will result in economic harm for its competitor and a change in the relative costs borne by each party much more quickly. Third, the case is decided by one's own national authorities rather than before an international panel. However, as was true of the first option, this strategy is also subject to certain constraints, and therefore, potential disadvantages. The most important of these is that one must be able to prove all of the required legal elements, as set forth in the ADA, for an antidumping duty. Otherwise, one's own government will not be able to impose antidumping duties in line with its WTO obligations. In addition, the domestic market must be large enough and its competitor's share within this market must be significant enough for the retaliatory antidumping duties to be sufficiently costly for one's competitor. Under this option (and unlike a successful outcome in the first option), the antidumping duties remain in place in the export market.

Nuctech decided to proceed with the second option. We do not know the extent to which the first option was also considered but it might not have been because there was not an apparent WTO violation by the Commission. Pursuing an antidumping case against Smiths might also be a preferable option for the Chinese government if support for Nuctech was indeed part of a wider industrial policy seeking to expand domestic production in this relatively high value-added sector.

Nuctech filed a petition with MOFCOM for an antidumping investigation on x-ray scanners from the European Union on 28 August 2009, some seven months after the EU's own antidumping investigation. The relevant products were within the Chinese HS category 9022.19.10. MOFCOM determined that the period of investigation for dumping would be 1 January 2008 to 31 December 2008 while the POI for injury to the domestic industry making the like product would be 1 January 2006 to 31 December 2008. The only exporter investigated in the dumping case was Smiths. Smiths provided responses to MOFCOM's questionnaires that were used to determine the possible dumping margins.

The investigated products in China–X-ray Scanners included non-medical x-ray scanning equipment. Note that the products covered in the two antidumping cases are slightly different. The Chinese dumping notice stipulates that the product included systems using more than 100 KeV while the EU announcement focuses on products with more than 250 KeV. This distinction will prove to be very important for the WTO litigation to come.

On 9 June 2010, MOFCOM announced a provisional dumping margin of 48.2% for Smiths. The pronouncement came only five days before the final antidumping duty of 34% announced by the European Commission in its own investigation of Nuctech. Seven months later, on 23 January 2011, MOFCOM announced a final antidumping duty of 33.5% for Smiths.Footnote ³³

What should we make of these patterns of counter-investigation and the nearly identical final duties imposed by China? Again, the process is foremost firm-driven. The move to bring antidumping duties within the arsenal of tools used to support Nuctech is one that was initiated by Nuctech itself. Faced with the prospect of being hurt by antidumping duties initiated by Smiths in the EU, Nuctech decided to fight back by petitioning to have similar duties imposed against Smiths in China.

While the actions in both Brussels and Beijing were clearly firm-driven, a question remains as to the role of the state in these actions. Were they simply passive actors? Did they take action only because they are required to do so in response to a domestic firm's petition for antidumping cases, as set forth in their own domestic laws and regulations? There are some hints to suggest that this indeed may have been the case. Nuctech relied upon the same Chinese law firm to defend its European and Chinese antidumping cases. When others have criticized the Chinese government for undertaking retaliatory antidumping cases, the offensive cases generally have been handled by one of the two dominant petitioners' law firms in China and not the law firm that handled Nuctech's cases. The fact that the follow-up case was not organized through the usual pattern may suggest that Nuctech (and/or its corporate law firm) was the key driver of the retaliatory action in China rather than the Chinese government.

However, given the close links between Nuctech and the government, it is difficult to rule out entirely the possibility that the Chinese government actively participated in the decision to file the antidumping case. If true, this pattern of government-sponsored tit-for-tat action is certainly not consistent with the spirit of the WTO system in which antidumping duties are to be used only in response to ‘unfair’ trading practices and not as a retaliatory measure against another country's trade action. To the extent that one has a problem with another country's trade actions, the WTO encourages the parties to resolve it diplomatically through consultations and turn to the Dispute Settlement Body if these consultations do not resolve the issue.Footnote ³⁴

Of course, we do not mean to necessarily single out China for potential bad behavior. As commentators have noted in Bown (Reference Bown2011), several WTO members have engaged in tit-for-tat retaliatory trade actions in response to trade remedies. To the extent that the actions are government-driven, the Chinese government is simply behaving in a similar manner to other major trading powers. Moreover, we cannot be certain that the actions here are government-sponsored. All we can note is that the firm that brought forward the action had close ties to the government and appears to have been supported through active strategic industrial policy. Most likely, both the firm and the government played active roles in triggering the problematic behavior that resulted in the WTO case – with the exact distribution between the two difficult for any outsider to discern.

Nevertheless, the overall point to note here is that the patterns of behavior are very different than those posited in the conventional account. Antidumping duties were not originally part of China's strategic industrial policy for championing a domestic firm in the x-ray scanners market. Antidumping duties probably entered into the Chinese government's calculus of how best to support Nuctech only after the EU imposed antidumping duties. The incorporation of antidumping duties into a government's industrial policy therefore may be a reactive development rather than one proactively crafted by economic policymakers. Furthermore, the actions of domestic firms and foreign governments may play a more important role in triggering this development than suggested in the existing literature.

This case study therefore highlights the fact that the conventional account is not the only path by which antidumping measures become enmeshed with a strategic industrial policy. The pathways are more varied. It is not always the case that the state simply takes the lead in using antidumping duties to protect an infant industry or salvage a dying one from import competition. This is even the case in China, where the state plays a relatively proactive role in the economy. Firms themselves may take the lead in introducing antidumping measures into the mix. This may be in reaction to another country's industrial policy or as a tit-for-tat response to another country's actions against it. Altogether, the dynamic between states and firms in drawing the two elements of antidumping and industrial policy into play is much more complex than has been acknowledged to date.

4.1 The EU's complaint

Rather than pressure the EU to ratchet up tensions even further with yet another antidumping measure against a Chinese product, Smiths opted instead to push the EU to file a WTO complaint against China over MOFCOM's antidumping order. The Commission clearly agreed that there was sufficient evidence that MOFCOM had committed legal errors when undertaking its antidumping investigation that would cause China to be in violation of its WTO commitments. As we shall see below, the EU found numerous bases for alleged violations.

The EU requested consultations over China's antidumping duties against European x-ray security inspection equipment in late July 2011, six months after the final Chinese antidumping determination. Consultations were held in mid-September 2011 and again in mid-October 2011 to no avail. On 8 December 2011, the EU requested that a Panel be established to adjudicate the dispute.

The EU complaint alleged violations of several provisions of the ADA, notably Articles 3.1, 3.2, 3.4, 6.2, 6.4, 6.5.1, 6.9, and 12.2.2. To simplify the discussion of the legal issues at stake, we group the EU's allegations into three categories of questions:

1. 1. Did Chinese authorities properly perform the injury analysis required for an antidumping investigation, including inter alia appropriately defining the boundaries of the product scope of the investigated sector when undertaking its price effects analysis and its analysis of the state of the domestic industry?

2. 2. Did Chinese authorities properly perform the causation analysis required for an antidumping investigation, including properly examining the causal relationship between dumped imports and injury and analyzing other known factors in its non-attribution analysis?

3. 3. Did Chinese authorities fulfill various WTO requirements on transparency and due process in its handling of the antidumping case, including requirements concerning the treatment of non-confidential information, the disclosure of ‘essential facts’, and the publication of methodologies used in the investigation?

Before discussing how the Panel addressed these issues in particular, we wish to highlight how this case fits with the broader trend of WTO litigation concerning China. China has been a longstanding target of other countries' antidumping measures, accounting for a disproportionate share of global antidumping actions.Footnote ³⁵ Not surprisingly, China has filed many complaints at the WTO over foreign antidumping actions (and, more recently, countervailing duties) against Chinese exports.Footnote ³⁶ In fact, seven of China's 12 WTO dispute cases as of 2014 have involved these ‘unfair’ trade restrictions.

In contrast, China is a relative newcomer among major trading countries in applying antidumping duties; however, in the past 15 years, it has emerged as a power in its own right.Footnote ³⁷ Through 2009, the actions of Chinese authorities on antidumping investigations went unchallenged at the WTO. None of the first 17 WTO complaints filed against China concerned antidumping. However, this pattern has shifted remarkably in recent years. Of the 14 complaints filed against China between 2010 and 2014, half have concerned China's antidumping regime.Footnote ³⁸ This case falls into this trend.

As mentioned, the Panel Report in China–X-Ray Equipment represents only the second time that the WTO dispute settlement body has issued a formal ruling on Chinese antidumping procedures. We expect several more WTO challenges and rulings in the near future given the importance of the Chinese market and the increasing willingness of China's trading partners to challenge its antidumping practices before the WTO.Footnote ³⁹ Just as American and European antidumping practices have been shaped through an iterative discourse with the WTO, so too will China's in the coming years.

On 26 February 2013, the Panel circulated its final report in the dispute. The EU prevailed on most, but not all of its claims against China. In the following subsection, we discuss the Panel's rulings on the three major sets of claims raised by the EU.

4.2 Injury analysis – ADA Articles 3.1, 3.2, and 3.4

The EU claimed that MOFCOM, when examining price undercutting and price suppressionFootnote ⁴⁰ as part of its injury analysis, did not base its findings ‘on an objective examination of positive evidence’ and consequently, China had violated its obligations under Articles 3.1 and 3.2 of the ADA.Footnote ⁴¹ At the heart of this debate is the question of how an investigating authority is to segment the product market/industry when performing the price analysis necessary to determine injury.

MOFCOM made its finding of price undercutting and price suppression by comparing the annual weighted average unit prices of all dumped imports with the annual weighted average unit process of all domestic sales.Footnote ⁴² In examining whether Nuctech suffered injury as a result of Smiths' actions, MOFCOM did not request Nuctech to provide transaction-specific data or data on a model-by-model basis.Footnote ⁴³ The EU noted that there are, however, important differences between ‘low energy’ and ‘high energy’ scanners, and that MOFCOM erred by not taking these differences into account.Footnote ⁴⁴ In particular, the EU noted that there were no exports of ‘high energy’ scanners from the EU during the period of investigation.Footnote ⁴⁵ China's failure to take this fact into account, the EU contended, meant that the price analysis methodology undertaken by MOFCOM was flawed since inadequate steps were taken to ensure that the prices compared were in fact comparable.Footnote ⁴⁶

China argued that MOFCOM did consider Smiths' arguments concerning ‘high energy’ and ‘low energy’ scanners and concluded that they were ‘like’ products.Footnote ⁴⁷ It cited the prior WTO ruling in EC–Salmon (Norway) Footnote ⁴⁸ noting that so long as an investigating authority finds the broad basket of goods under investigation and the broad basket of domestic goods to be ‘like’, this is sufficient; it need not be the case that each good within the basket of domestic goods is ‘like’ each of the products under investigation.Footnote ⁴⁹ Furthermore, it noted that Smiths had only raised the issue of comparability in the context of the scope of products under investigation, but not as part of the price undercutting or suppression analysis.Footnote ⁵⁰ Nor had Smiths provided the data necessary for MOFCOM to undertake a model-by-model price comparison. Thus, China argued that given the circumstances, MOFCOM's use of the weighted average methodology was reasonable and appropriate.Footnote ⁵¹

The Panel rejected China's arguments and sided with the EU. It did not accept China's argument that its approach was reasonable given the circumstances, noting that even if Smiths had provided such data, it would have been impossible for MOFCOM to perform the proper analysis because it had not requested such pricing data from Nuctech.Footnote ⁵² Because MOFCOM had failed to demonstrate its willingness to undertake the proper methodological adjustments (regardless of whether data were forthcoming or not), the Panel concluded that ‘MOFCOM's price undercutting and price suppression analyses … were not based on an objective examination of positive evidence’.Footnote ⁵³ Consequently, the Panel ruled that ‘China acted inconsistently with Articles 3.1 and 3.2 of the [ADA] by failing to ensure that the prices it was comparing as part of its price effects analysis were actually comparable’.Footnote ⁵⁴

In addition, the EU asserted that China had further erred in its injury analysis because MOFCOM did not base its finding on positive evidence about the state of the domestic industry.Footnote ⁵⁵ Consequently, the EU alleged that China also violated its obligations under Articles 3.1 and 3.4 of the ADA. In particular, the EU drew attention to discrepancies between the data cited by MOFCOM and (1) the data supplied by Nuctech in its petition and questionnaire, and (2) the data in publicly available documents, such as Nuctech's parent company annual reports and agency filings.Footnote ⁵⁶

The Panel sided with China on this issue. While acknowledging that there were indeed discrepancies, the Panel noted that the relevant question was ‘whether the European Union has established a prima facie case that the evidence upon which MOFCOM relied in making its injury determination was not of an affirmative, objective and verifiable character and was not credible’.Footnote ⁵⁷ The Panel held that the EU had failed to present adequate evidence to make this case. In particular, the Panel noted that it was not convinced by the EU's suggestion that the provided data could not have been modified properly, and that the EU had not demonstrated that the data, assuming they were modified, were not credible.Footnote ⁵⁸

The EU further alleged that China had acted inconsistently with ADA Articles 3.1 and 3.4 because MOFCOM did not examine all of the factors listed in Article 3.4 as required for the injury analysis.Footnote ⁵⁹ In particular, the EU drew attention to the fact that MOFCOM had failed to examine ‘the magnitude of the margin of dumping’.Footnote ⁶⁰ In response, China noted that MOFCOM had listed the magnitude of the margins of dumping for Smiths and other producers in two sections of its final determination, thereby suggesting that MOFCOM had indeed considered this factor.Footnote ⁶¹ The Panel rejected China's argument by noting that a mere listing of the facts was insufficient evidence of consideration of the factor by the investigating authority.Footnote ⁶² Therefore, the Panel also concluded that China had acted inconsistently with ADA Articles 3.1 and 3.4 because MOFCOM failed to demonstrate that it had evaluated the magnitude of the margins of dumping in its injury analysis.Footnote ⁶³

Finally, the EU noted that MOFCOM failed to provide a compelling explanation for why it found injury to the domestic industry when 9 of the 16 indicia of the state of the industry were found to be positive rather than negative.Footnote ⁶⁴ The Panel agreed with the EU that MOFCOM's treatment of these 16 factors ‘did not reflect an objective examination of the evidence’.Footnote ⁶⁵ Consequently, it decided that China had also violated its obligations under Articles 3.1 and 3.4 of the ADA on this basis.Footnote ⁶⁶ The Panel's view was that rather than simply listing the factors and the trends observed, MOFCOM should have explicitly analyzed each of the 16 factors and assigned them appropriate weights in its overall injury analysis.Footnote ⁶⁷ The Panel held that MOFCOM's error was one of methodology rather than substantive explanation; the Panel did not think it necessary to decide whether MOFCOM was obliged to provide a more compelling explanation than that which had been given.Footnote ⁶⁸

To summarize, with respect to the injury analysis, the Panel found that several of MOFCOM's actions were inconsistent with China's treaty obligations as set forth in the ADA. The most significant findings were those concerning the industry definition used by MOFCOM. We will return to examine this issue in our analysis in section 5.

4.3 Causation analysis – ADA Articles 3.1 and 3.5

The EU also argued that China violated Articles 3.1 and 3.5 of the ADA ‘because MOFCOM failed to make an objective determination, on the basis of all relevant evidence before it that the dumped imports were, through the effects of dumping, causing injury’.Footnote ⁶⁹ In other words, the EU suggested that shortcomings in MOFCOM's injury analysis also led it to perform an improper causation analysis.

The Panel ruled that MOFCOM had erred in its price analyses because it failed to take into account differences across the products (i.e., between ‘high energy’ and ‘low energy’ scanners), and the Panel agreed that ‘flaws in the price effects analysis also undermine MOFCOM's conclusion on the causal link between the subject imports and the injury suffered by the industry’.Footnote ⁷⁰

Again, the issue of industry definition and proper product differentiation proves fatal for China. Not only did MOFCOM's actions here cause China to be in violation of its WTO obligations concerning the ADA's requirements for the injury analysis, but they also led to violations of China's obligations concerning how the causation analysis should be performed.

The Panel also found MOFCOM's conclusion that dumped imports were the only cause of injury over the period of investigation to be problematic. Given that the price of the domestic good continued to be lower than that of imports, the Panel asked why couldn't Nuctech simply increase its prices at least to the level of Smiths' prices.Footnote ⁷¹ Was it the case that other factors, besides dumping, were also at work in causing Nuctech's injury? The fact that MOFCOM did not address this issue in its final determination, the Panel found, suggested that MOFCOM ‘did not conduct an objective examination of the evidence’ in violation of ADA Articles 3.1 and 3.5.Footnote ⁷²

In addition, the EU also argued that MOFCOM failed to consider the relevance of other known factors as part of its non-attribution analysis. WTO law requires that, in the course of determining whether dumped imports are causing injury to the domestic industry, the investigating authority is obligated to ‘examine any known factors other than the dumped imports which at the same time are injuring the domestic industry’.Footnote ⁷³ The EU alleged that MOFCOM ignored five known factors raised by interested parties, ‘namely (i) the impact of the global financial crisis; (ii) fair competition between Nuctech and other producers; (iii) Nuctech's aggressive business expansion; (iv) Nuctech's aggressive pricing policy; and (v) Nuctech's start-up situation’.Footnote ⁷⁴ China, in turn, argued that before MOFCOM was obliged to consider any of these known factors, Smiths was obliged to present evidence that these factors were causing injury to Nuctech and had failed to do so.Footnote ⁷⁵

With respect to the first and last of the five factors put forward by the EU, the Panel sided with China. It noted that MOFCOM's handling of the factors was sufficient given the arguments put forward by Smiths.Footnote ⁷⁶ The Panel sided with the EU on the three remaining issues. It concluded that ‘MOFCOM did not adequately engage with the evidence and arguments submitted by Smiths to the effect that Nuctech could not offer the same products as Smiths, or could not offer products of the same quality’.Footnote ⁷⁷ The Panel further determined that MOFCOM should have also considered the impact of Nuctech's capacity expansion and its rising inventories,Footnote ⁷⁸ as well as investigated further the possibility that Nuctech was pursuing an aggressive pricing strategy as alleged by Smiths.Footnote ⁷⁹

Overall, although the Panel did not agree with all of the EU's arguments concerning causation-related claims, it found several of them to be valid. Specifically, the Panel specified three shortcomings: (1) MOFCOM had relied on a WTO-inconsistent price effects analysis and injury findings for its causation analysis; (2) it ‘failed to consider certain “known factors”’; and (3) it ‘failed to consider evidence relating to other factors that it did explicitly consider, in its non-attribution analysis’.Footnote ⁸⁰ Consequently, the Panel concluded that China had acted inconsistently with its WTO requirements, as set forth in ADA Articles 3.1 and 3.5, in determining that Smiths' dumping had caused injury to Nuctech.

4.4 Issues concerning transparency and due process – ADA Articles 6.5, 6.9, and 12.2

The EU also raised a series of issues alleging procedural shortcomings by MOFCOM. First, the EU challenged MOFCOM's handling of non-confidential information submitted by Nuctech. A petitioner in an antidumping case often must submit information that it considers confidential. To ensure that the respondent has a chance to file an effective rebuttal, WTO law requires that ‘[t]he authorities shall require interested parties providing confidential information to furnish non-confidential summaries … in sufficient detail to permit a reasonable understanding of the substance of the information submitted in confidence’.Footnote ⁸¹ The law contains an exception for ‘exceptional circumstances’ in which ‘such information is not susceptible of summary’.Footnote ⁸² When invoked, the authorities must provide a statement explaining why summarization is not possible.Footnote ⁸³

Smiths exported four models of x-ray scanners to China. In the non-confidential summary provided to Smiths, MOFCOM had allowed Nuctech to designate the models that Nuctech alleged were being dumped as simply ‘Model 1’ and ‘Model 2’. China contended that Smiths, on the basis of the information provided, should have been able to figure out the two models being referenced since Smiths exported only four models.Footnote ⁸⁴ The EU, however, argued that this labeling failed to meet the requirements of Article 6.5.1.Footnote ⁸⁵ The Panel agreed with the EU, noting that it was the investigating authority's responsibility to require ‘the interested party furnishing the information to provide a summary thereof, rather than requiring the other party to infer, derive, or piece together a possible summary of the confidential information’.Footnote ⁸⁶ The Panel also upheld the EU's claim that the non-confidential summaries provided by Nuctech of relevant Exhibits and responses to questions were also inadequate.Footnote ⁸⁷

The EU also raised an objection to MOFCOM's failure to provide a summary of certain confidential information submitted by the Public Security Bureau (PSB) of the Civil Aviation Administration. China claimed that MOFCOM had received an oral explanation from the PSB about why its information could not be summarized and argued that it did not provide this explanation because the reasons themselves were confidential.Footnote ⁸⁸ The Panel rejected China's argument, stating that China nevertheless should have explicitly invoked the ‘exceptional circumstances’ exception and provided a ‘statement of the reasons why summarization of the underlying confidential information was not possible’.Footnote ⁸⁹

Second, the EU claimed that MOFCOM failed to disclose certain essential facts that formed the basis of its decision to apply definitive antidumping duties.Footnote ⁹⁰ The EU accused MOFCOM of inadequate disclosure in four instances.Footnote ⁹¹ The Panel accepted the EU's arguments in three of the four instances.Footnote ⁹² In two of these instances, the Panel decided that MOFCOM was not obliged to disclose as much as the EU had contended, but that MOFCOM's disclosures were nevertheless inadequate.Footnote ⁹³ The Panel consequently held that China was in violation of its obligations under Article 6.9 of the ADA.

Finally, the EU challenged China on the basis of shortcomings concerning the public notice provided by government authorities. The chapeau of Article 12.2 of the ADA provides that the government, when issuing any preliminary or final determination, ‘shall set forth … in sufficient detail the findings and conclusions reached on all issues of fact and law considered material by the investigating authorities’. Furthermore, the first sentence of ADA Article 12.2.2 provides that the ‘public notice of conclusion … of an investigation in the case of an affirmative determination providing for the imposition of a definitive duty … shall contain, or otherwise make available through a separate report, all relevant information on the matters of fact and law and reasons which have led to the imposition of final measures'. The second sentence then provides that ‘the notice or report shall contain … reasons for the acceptance or rejection of relevant arguments or claims made by exporters or importers'.

The EU made two sets of claims regarding Chinese violations of Article 12.2.2. The EU claimed that MOFCOM's public notice failed to include relevant information concerning the price effects analysis, the margin of dumping established for Smiths, and the residual antidumping duty.Footnote ⁹⁴ The Panel found that MOFCOM had failed to ‘provide relevant information regarding its price effects analysis and its determination of the residual rate’.Footnote ⁹⁵ However, it rejected the EU's claims that calculations of the dumping margin and residual rate must be included in the public notice.Footnote ⁹⁶ On this last question, the Panel's holding falls squarely in line with the prior one established in the China–GOES case.Footnote ⁹⁷

The EU also alleged that MOFCOM violated Article 12.2.2 of the ADA by not explaining in its public notice why it rejected arguments made by Smiths concerning three different issues.Footnote ⁹⁸ The second sentence of Article 12.2.2 requires that the investigating authority include its rationale for acceptance or rejection of the parties' arguments in its public notice, and the Panel emphasized that ‘it is particularly important that the “reasons” … be set forth in sufficient detail to understand why their arguments or claims were treated as they were, and to assess whether or not the investigating authority's treatment of the relevant issue was consistent with domestic law and/or the WTO Agreement’.Footnote ⁹⁹ The Panel then found that MOFCOM violated this requirement in one of the three instances, where MOFCOM never addressed an argument raised by Smiths concerning the arms-length nature of its sales to affiliated distributors in its final determination.Footnote ¹⁰⁰ However, the Panel declared that with respect to arguments raised in the other two instances, either the MOFCOM's explanation was sufficientFootnote ¹⁰¹ or the EU had failed to establish a prima facie case for its claim.Footnote ¹⁰²

Overall, the EU successfully prevailed in a number of transparency-related claims against China. Ranging from the treatment of non-confidential summaries to the completeness of the public notice provided, the WTO Panel Report draws attention to deficiencies in the due process and other procedural protections afforded in administrative litigation before Chinese government authorities. These may prove relatively straightforward for MOFCOM to fix, but in the meantime, they pose two consequences. The first is that China will need to undo its antidumping duties against Smiths while MOFCOM engages in these procedural fixes. The second is that future respondents before MOFCOM will have access to greater information – an implication that we will explore in greater detail in the next section.

4.5 Epilogue

As the above case study demonstrates, WTO litigation provides a means for a WTO member to alter the impact of antidumping duties used to support (or combat) an exporter supported by industrial policy. For this strategy to be effective, however, the WTO member must be able to allege that its trading partner committed certain WTO-inconsistent acts when imposing the antidumping duties. Inexperienced investigating authorities more commonly commit such acts, although they are by no means the only ones.Footnote ¹⁰³ The EU, in this instance, successfully challenged several of MOFCOM's acts. China, by contrast, had chosen earlier not to challenge any of the Commission's acts.

China also decided not to appeal the Panel Report. The Dispute Settlement Body adopted it on 24 April 2013. Ten months later, on 26 February 2014, MOFCOM terminated the antidumping duty on Smiths x-ray scanners, thereby bringing China into full compliance with the Panel ruling. Table 1 provides a summary of the full timeline of the two antidumping cases as well as the WTO case.

Table 1. Timeline of Relevant X-ray Scanner Antidumping Cases between the EU and PRC

WTO litigation therefore serves as a mechanism to dramatically alter the payoffs from employing antidumping duties in support of an industrial policy. Although the issues litigated are highly technical, at the end of the day the litigation result provided Smiths with a clear advantage. Without the WTO case, both Smiths and Nuctech would continue to face temporary trade barriers in a key export market. The tit-for-tat antidumping measures seek to neutralize each other out, thereby allowing any industrial policy at work to continue to exert its impact on the global market. The WTO ruling, however, changes this dynamic. As of now, only the EU continues to maintain antidumping duties in an effort to thwart Nuctech from gaining a greater foothold in Smiths' home market. To the extent that the Chinese government may be relying upon industrial policy instruments to support Nuctech, that arsenal does not, at present, include antidumping duties. WTO litigation proved to be an effective means for the EU to undo Chinese efforts to retaliate against Smiths and reduce the payoffs from a Chinese industrial policy.

However, the saga may not have come to an end. MOFCOM has announced that it plans to reopen an investigation on x-ray equipment from the EU. The prospect of continuing tit-for-tat antidumping sanctions between the EU and China lingers on the horizon.

What the China–X-Ray Equipment case study demonstrates, therefore, is that the relationship between antidumping duties and industrial policy is much more complex than that posited in the conventional account. A back-and-forth dynamic exists between states and firms, and between litigation before domestic administrative agencies and the WTO. Minutiae of WTO law, as discussed in this section, can factor in this dynamic, as each side seeks to gain a relative advantage over the other in the quest for a growing global market.

5.1 Are antidumping measures an effective supplemental instrument for strategic industrial policy? Evidence from the Nuctech–Smiths dispute

We have suggested that the Chinese government may be using trade remedy laws to support a broader industrial policy goal for the x-ray scanner market, though it is unclear whether this is in reaction to a firm-led initiative or part of a direct partnership between the government and firm. Regardless, what is clear is that the Chinese government took actions in contravention of its WTO obligations that benefited its domestic producer through antidumping duties. What triggered this action was the need to react forcefully to its trading partner's antidumping measures designed to prevent a Chinese upstart from expanding into an established high-tech market.

Can such a strategy work? Past studies have shown that by raising tariffs, antidumping duties can be an effective tool in reducing imports from a specific foreign competitor.Footnote ¹⁰⁴ They are more effective in helping the threatened domestic industry when there are no alternative foreign sources easily available to simply replace the targeted imports. But can they also be an effective tool at countering the antidumping duties that a trading partner has imposed against a domestic producer supported through a strategic industrial policy? If so, one should see evidence in trade flows of x-ray scanners between the EU and China.

One of the case's complications is that a number of potentially disparate products fall within this HS category.Footnote ¹⁰⁵ This is in itself not unusual in antidumping cases. Many countries investigate imports based on detailed product descriptions that do not necessarily line up one-to-one with a single HS code.

In this particular instance, the Chinese eight-digit HS category (9022.19.10) identified in the investigation includes not only products such as security equipment to examine large truck containers and railway cars but also airport hand luggage scanners. Whether these are a single ‘like’ product is at the heart of the WTO dispute, which will be discussed in more detail in the next subsection.

Nevertheless, we can still examine the relevant data to discern what impact antidumping duties had on trade flows between the EU and China in the relevant eight-digit product categories in each market.Footnote ¹⁰⁶Table 2 shows that in the years preceding the antidumping investigations in each market, both the EU and China were expanding their exports of x-ray scanners into the other's market.

Table 2. EU and China trade in non-medical x-ray equipment

Notes: * Source: UN Comtrade, value in thousands of US$

** Source: Eurostat, value in thousands of euros

Let us first examine the inroads made by Nuctech into the European market. Official European data for HS category 9022.19.00 (the most relevant individual category for the case given the EU's eight-digit coding) shows a dramatic increase in Chinese exports: a near tripling of import value from €3.96 million in 2005 to €11.87 million in 2008, the year before the antidumping case was initiated in the EU.Footnote ¹⁰⁷ Nuctech's success in Europe prompted Smiths to file its antidumping case, triggering the chain of events at issue in this case.

While Nuctech was making progress penetrating Smiths' once dominant hold of the European market, Smiths was not simply standing still. It is important to note this is not the story of yet another once-dominant Western firm simply being pummeled globally by a new low-cost Chinese entrant supported by industrial policy. In the years prior to the anti-dumping case, Smiths too was making inroads into the Chinese market that Nuctech dominated. UN Comtrade statistics for China show that the total import value from the EU for the eight-digit product category (HS 9022.19.10) doubled from $13.1 million in 2005 to $26.6 million in 2008, almost all of which came from Germany (the production home of the x-ray scanners named in the petition).

The data reported in the preliminary Chinese antidumping decision provide additional details for the more narrow set of investigated products within the eight-digit HS category. It notes that investigated EU imports from Germany rose 87% from 179 units in 2006 to 336 units in 2008.Footnote ¹⁰⁸ It also reports that EU imports' share of the Chinese market for the investigated products rose from 11% in 2006 to 16% in 2008.Footnote ¹⁰⁹ The relatively modest increase in market share, despite the significant increase in the volume of European imports, suggests that overall Chinese demand for these products must have been increasing fairly rapidly in this period. This should not be surprising, given China's heavy infrastructure spending in the mid to late 2000s.

The preliminary Chinese antidumping decision also notes some import pricing data for the period of investigation that were important in the Panel's decision. The average unit value reported by MOFCOM for investigated EU products rose in every year of the investigation: RMB 463,520 in 2006 to RMB 468,313 in 2007 to RMB 504,767 in 2008.Footnote ¹¹⁰ In contrast, the report notes a fall in domestic prices of the similar goods of 49% from 2006 to 2007 and a further decrease of 47% in 2008.Footnote ¹¹¹

In short, UN Comtrade data and China's own report suggest the following important patterns for the three years prior to the antidumping investigation: (1) the price (unit value) of EU x-ray scanners was increasing in China; (2) Chinese prices (unit value) for the x-ray scanners were falling; and (3) EU exports to China nonetheless were increasing. (Note that this combination of factors lies at the heart of the arguments about whether EU imports could be causing injury to Chinese products.)

What we see from the trade flow statistics in the years prior to the antidumping investigation is that both European and Chinese producers appeared to be benefiting from expanded opportunities created by trade. At the same time that exports of European x-ray scanners to China were increasing dramatically, Chinese x-ray scanners were also making strong progress into the European market. Table 2 shows that between 2005 and 2008, EU exports to China of the relevant product category expanded by $13.5 million while Chinese exports to the EU expanded by €7.9 million.Footnote ¹¹²

Nevertheless, this is clearly not how Smiths saw the situation. Despite the fact that its exports to China were increasing dramatically, it was concerned about the threat it faced from Nuctech in its home market. Smiths was determined to lessen pressure from its Chinese competitor by filing the antidumping action against Nuctech in the EU.

The efforts by Smiths and the Commission to restrict Nuctech's exports to the EU followed a period when the Chinese market was relatively open to EU x-ray scanner exports. Chinese government agencies had procured significant volumes of x-ray scanning equipment from the EU in the mid-2000s. Although China may have executed an industrial policy to expand domestic production in this sector, it had not completely shut its market to foreign firms. Instead, it had signed contracts with foreign incumbents, even as it worked to provide expanded opportunities for the upstart domestic firm. But if the Europeans were now going to try to shut Nuctech out of its market using antidumping duties, why not do the same in response?

What impact then did the antidumping duties imposed on both parties have on trade flows? As expected, imports from the affected party dropped dramatically following the imposition of antidumping duties. Chinese exports of x-ray scanners to the EU in 2010 (i.e., the first full year following the imposition of provisional antidumping duties by the Commission) fell by 62% from their peak level in 2008, the last full year prior to the antidumping investigation. Similarly, European exports of x-ray scanners to China in 2011 (i.e., the first full year following the imposition of provisional antidumping duties by MOFCOM) fell by 87% from their peak level in 2008. In short, the trade statistics suggest that the antidumping duties led to significantly lower imports.

If Nuctech and the Chinese government's aim was to punish Smiths and the EU for hitting Nuctech with antidumping duties, then the retaliatory tit-for-tat measures clearly worked. China demonstrated clearly that it could give as good as it could get. Trying to shut out a rising Chinese upstart, especially one supported by Chinese industrial policy, was not without consequences for the established incumbents.

There is, of course, another question: Were the tit-for-tat actions effective in bolstering the domestic firm's competitiveness in the aftermath of the antidumping measures imposed against it overseas? Did the Chinese government provide its upstart producer with a boost by closing off more of the domestic market as a response to the Smiths' decision to press the EU for antidumping duties? Again, the trade flow data suggests the possibility that the answer may be yes. Chinese exports in this product category fell in 2009 and 2010 subsequent to imposition of antidumping duties. However, Chinese exports to the EU have rebounded significantly, and in 2013, they reached 70% of their value prior to the antidumping duties.

These results suggest that if a government is seeking to counter the negative impact of an antidumping action imposed by a trading partner on its producer, a retaliatory tit-for-tat antidumping measure in the same market can ‘work’. Retaliation is effective in the sense that it can alter patterns of trade positively, at least in the near term, for the domestic firm. However, it is important to note two caveats: First, the likelihood of success associated with such a strategy may depend, among other things, on the size of the domestic market of the upstart. If it is significantly large, as in the case of China, then additional protection afforded by the tit-for-tat measure can provide a boost to offset the harm suffered from the antidumping measure imposed overseas by the incumbent's government. Second, depending on the market, one need not turn necessarily to antidumping measures as a means of retaliation and assistance to the upstart domestic firm. Other policy instruments, such as government procurement, may also exist, especially for sectors heavily reliant on government contracts and for governments that are not signatories to the GPA. In that sense, this may again be further evidence that the process here was at least partially firm-driven rather than purely government-driven.

5.2 Price comparisons in an injury investigation: lingering questions left unresolved by the Panel Report

The China–X-Ray Equipment decision is also significant in its attempt to clarify open questions over how authorities in antidumping investigations are to perform price comparisons when assessing injury pursuant to Article 3 of the ADA. Below, we explain the significance of this issue and attempt to highlight its importance. We suggest that the Panel was correct to take steps to clarify the legal discipline, but we argue that it did not spell out enough. We end by highlighting several questions left unanswered by the Panel ruling.

The question of the scope of the product under investigation is a key element of any antidumping investigation. Investigators must resolve this question before they can assess whether dumping is occurring and whether it is causing injury to the domestic producer to require the imposition of antidumping duties. Their decision, in turn, will affect several elements of the investigation, including what data are collected from which set of firms, foreign and domestic.

The ADA provides inadequate guidance on how investigating authorities are to define ‘like’ products. ADA Article 2.6 simply states that the ‘like product shall be interpreted to mean a product that is identical, i.e., alike in all respects to the product under consideration …’ WTO Panels have afforded investigating authorities broad discretion, but as the Panel ruling in China–X-Ray Equipment makes clear, there are limits to this discretion.

During the investigation, Smiths and MOFCOM vehemently disagreed over the proper product scope for the investigation. Smiths claimed that low-energy and high-energy scanners are distinct goods on account of differences in their physical characteristics, end-uses, technological features, mechanical features, manufacturing processes, and prices.Footnote ¹¹³ Furthermore, Smiths asserted that the two do not compete against each other. As Smiths exported only low-energy scanners during the period of investigation, Smiths argued that high-energy scanners should be excluded from the product scope under investigation.Footnote ¹¹⁴ MOFCOM, on the other hand, disagreed, and considered both high- and low-energy scanners as part of its investigation, which contributed to its finding of injury to Nuctech on account of price undercutting and suppression from Smiths' imports.

The energy level of scanners is measured in kiloelectronvolts (keV). Low-energy scanners, for example, are the ones used to scan luggage at an airport or bags at a subway station. High-energy scanners, on the other hand, are the ones used to scan railcars and trucks. The prices between low- and high-energy scanners differ greatly because of complexity and size. The EU maintained that these are two distinct categories; a consumer for one would never consider the other when making its purchases. China, on the other hand, argued that ‘there is “no hard and clear dividing line” between high- and low-energy scanners, but rather a “single continuum”’ of scanners.Footnote ¹¹⁵

Note that this issue is by no means specific to x-ray scanners. Consider, for example, flat-screen monitors. Are there two distinct categories – small flat-screen monitors, for example, used with home computers, and large flat-screen monitors used to project information at airports? Or is there a single continuum of flat-screen monitors? In each of these instances, there are clear distinctions between the categories, including in terms of price and consumer end-use. But are the differences so distinct that they should not be considered ‘like’ products? If so, how does one draw the line between categories? For x-ray scanners, is it at 250 KeV, 300 KeV, or some other level? Similarly, for flat-screen monitors, do we divide the two categories at 16 inches, 30 inches, or some other level?

The question is to what extent decisions made on such questions by administrative agencies in domestic litigation should be subject to review by the WTO. In the past, the WTO has shown itself to be very reluctant to challenge the line-drawing determinations of domestic authorities. For example, in another recent case, EU–Footwear, the tables were turned, with China challenging the EU's determination in an antidumping case on Chinese footwear. The Commission had decided to lump adult dress shoes, children's leather sandals, and low-end athletic shoes into the scope of the product investigation, but not high-end athletic shoes. The dividing line, according to the Commission, was €7.50 per pair; special-technology athletic shoes above this price were excluded from the investigation whereas those below it were included.Footnote ¹¹⁶ China challenged the EU's line-drawing when determining the scope of products under investigation, but lost. The Panel reaffirmed earlier decisions suggesting that it did not have the authority to review the domestic administrative agency's determination of the product scope under investigation through ADA Article 2.6.Footnote ¹¹⁷

Note that from an economist's point-of-view, substitutability in consumption is a key factor in determining product grouping. Economic analysis would suggest that adult formal shoes and children's sandals are very distinct products and should not be grouped together. On the other hand, it is unclear that this is the case with high-end athletic shoes. An adult male would never consider children's footwear for a stroll around town but might choose athletic shoes. If the dividing line is drawn at €7.50 per pair for athletic shoes, then the burden would appear to fall on the authority drawing this line to show that a consumer would only substitute shoes below this line but not above it. Yet, that was not the approach embraced by the WTO in EU–Footwear. Instead, the earlier Panel left investigating authorities with wide discretion in determining ‘like’ products and in deciding where to draw the line in deciding to exclude a subcategory of the ‘like’ product.

While the China–X-Ray Equipment case also involves line drawing, we draw attention to the fact that it has the opposite fact pattern. In EU–Footwear, the litigant (China) asserted that the defendant (the EU) had improperly drawn a line to subdivide a product category, thereby causing the product scope under investigation to be too narrow and discriminating against its producers' interest. In contrast, in the case at hand, the litigant (the EU) asserted that the defendant (China) had improperly failed to draw a line to subdivide a product category, thereby causing the scope to be too wide and discriminating against its producers' interest.

Possibly on account of the lessons learned from China's failed litigation claims in EU–Footwear, the EU decided not to challenge China on MOFCOM's determination that low- and high-energy x-ray scanners are ‘like' product.Footnote ¹¹⁸ EU litigators did not argue, as Smiths had, that the two categories are distinct and MOFCOM was wrong to lump the two together. They avoided making any allegation that China violated ADA Article 2.6. Instead, the EU mounted a more indirect attack.

The EU noted that ‘there is a wide and recognisable gap between high- and low-energy scanners', leading to ‘very large disparities in price’.Footnote ¹¹⁹ However, it did not then proceed to argue that this gap should have been the basis for excluding high-energy scanners from the investigation altogether. Instead, the EU simply asserted that MOFCOM should have taken these price disparities into account when conducting its price undercutting and price suppression analyses and that its failure to do so violated Article 3.2 of the ADA.Footnote ¹²⁰

When it comes to the question of how investigating authorities are to compare prices during an injury analysis, the ADA is relatively silent. This stands in sharp contrast to the related question of how price comparisons are to be done as part of the calculation of the dumping margin. The second sentence of Article 3.2, which covers the injury analysis, simply states that authorities are to consider price undercutting and price suppression, without any mention of how this is to be done.Footnote ¹²¹ On the other hand, ADA Article 2.4.2, which pertains to the calculation of the dumping margin, sets forth in excruciating detail the permissible methodology for price comparisons.Footnote ¹²² The law's meaning has been further clarified through a long line of jurisprudence surrounding questions of zeroing.

Past WTO Panels, however, have made clear that the law on price comparisons for dumping margin calculations cannot be simply transposed to Article 3 and the injury analysis.Footnote ¹²³ However, the law on price comparisons for the injury analysis has not been well established by subsequent jurisprudence. In China–GOES, the Appellate Body simply agreed that the Panel below was correct in declaring that ‘an investigating authority must ensure comparability between prices that are being compared’ as part of the injury analysis.Footnote ¹²⁴ How specifically this is to be done was a question left for a future case.

China–X-Ray Equipment, therefore, should be understood as an attempt by a later Panel to clarify further the Appellate Body's ruling in China–GOES on ADA Article 3. While it does so to an extent, we argue that the Panel Report still leaves questions unresolved. We explain in greater detail below.

After determining that high- and low-energy scanners are ‘like’ products, MOFCOM proceeded to compare the prices of the dumped imports and the like domestic product for the purposes of its price undercutting and price suppression analysis. For both the dumped foreign imports and the like domestic products, MOFCOM calculated the average unit value of the product by dividing total value by total quantity.Footnote ¹²⁵ China argued that since MOFCOM applied the same methodology of calculating a weighted average for the dumped imports and the like domestic products, it fulfilled the requirements of ADA Article 3.2 which simply requires that such a comparison be made.Footnote ¹²⁶

The EU contended that MOFCOM's technique was insufficient to meet the Appellate Body's command in China–GOES that an investigating authority must ‘ensure comparability’ across products when undertaking a price undercutting analysis. The EU argued that even when investigating authorities find goods to be ‘like’ product, they must nevertheless inquire into whether price differences exist across the goods that fall within the category of ‘like’ products. Where these differences are significant enough, then the authorities must make adjustments to ensure proper comparability.Footnote ¹²⁷ In other words, authorities could not simply take weighted averages of the two bundles – dumped imports and like domestic products – without taking into consideration the product composition of the bundles, as MOFCOM had done. To do so violated WTO law.

Note that when viewed from a strict textualist perspective, the validity of the EU's argument is questionable. Article 3.2 of the ADA simply states that ‘the investigating authorities shall consider whether there has been a significant price undercutting by the dumped imports as compared with the price of a like product’. If A and B are ‘like’ products, then Article 3.2 suggests that investigating authorities are at liberty to compare the prices of A and B as they see fit. Nothing in the text itself suggests that under certain circumstances, the price of dumped imports of A cannot be compared with the price of domestically produced B, but must instead be compared with domestically produced A. The textualist would further note that the bare language of Article 3.2 stands in stark contrast to that of Article 2.4.2, which places further constraints to ensure price comparability. The fact that this detail was consciously omitted from Article 3 lends further support to an argument that authorities enjoy wide discretion in how they choose to compare the prices of A and B once they find that the two are ‘like’ products.

For example, consider if A is a red apple and B is a green apple. Is it the case that the price of a dumped red apple must be compared with the price of a domestically produced red apple? Or can it also be compared against the price of a domestically produced green apple? In this example, the price differences between the two are not usually that significant. To an outsider, the question might seem to be trivial, although to a fruit farmer with thin profit margins, the issue of a few cents may well matter. Red and green apples are clearly ‘like’ products, and Article 3.2 seems to grant authorities discretion to compare the prices of dumped red apples against domestic green apples.

But what if the price differences are much larger? Suppose A is a white mushroom and B is a truffle, which is much rarer and more expensive. Or A is farmed salmon and B is wild salmon. One might argue that the two are not ‘like’ products because of different end-uses. But suppose, as was true here, both sides agreed that the authorities enjoyed the discretion to find them to be ‘like’ products. Do they, because of price differences, no longer enjoy the flexibility to compare A versus B?

The cleaner approach to handling such cases would simply be to opine further regarding the ‘likeness’ jurisprudence in Article 2.6 of the ADA. But because such challenges have failed, WTO members are increasingly turning to claims under Article 3 instead, as the EU did in this case. The Appellate Body's ruling in China–GOES had already opened the door for adjudicators to suggest that investigating authorities do not enjoy unfettered discretion when comparing prices between foreign and domestic goods. The Panel in China–X-Ray Equipment goes a bit further to fill in some of the details.

The Panel's holding in this case should be understood as suggesting that even when A and B are determined to be ‘like’ products, the investigating authorities must consider the statistical distribution of A and B within the bundle of dumped imports and like domestic products when making its price comparisons under Article 3. It cannot simply just compare the two bundles without due regard to their composition.

Inclusion of different versions of a product is not necessarily inappropriate in calculating averages so long as the underlying statistical distribution is similar. Unfortunately, this was not the case in this particular investigation. In particular, the EU firm sold none of the high-energy scanners in China during the period of investigation while Nuctech produced both high- and low-energy scanners. If the difference in price across the two is not significant, as is the case with red and green apples, then the differences in composition may not matter much to the eventual outcome. But when they are significant, as is the case with x-ray scanners and even mushrooms, then failure to make adjustments will skew the outcome. Including the more expensive high-energy scanners in the determination of Chinese prices would result in oversampling of the high priced goods and increase the price-underselling margin.

To see why this is the case, consider the following simple example. Suppose that Smiths sells two low-energy scanners at $200 but sells no high-energy scanners. Nuctech also sells a single low-energy scanner at $200 plus a high-energy version at $1,000. Under the MOFCOM procedures used in this case, the foreign average unit value is $200 while the domestic average is $600, indicating significant price underselling by Smiths. But this comparison is only appropriate if the two scanners are ‘similar’. If the authorities consider low-end scanners as the correct comparison, there is no price underselling at all since both foreign and domestic sales are an average of $200 per unit.

MOFCOM defended its approach by arguing that the procedure is fair given that the same procedure is used for both domestic and foreign prices. In particular, MOFCOM calculated unit values for the EU firms by dividing the total value of imports (from customs data) by the number of units sold. They used aggregate sales data from Nuctech and also divided the value of domestic sales by the number of units sold. This approach is particularly problematic when individual prices vary significantly across the data sample: simple averages can hide enormous variability within. Left unsaid by MOFCOM is that Nuctech's exports of the relatively high-energy scanners (i.e., greater than 250 KeV) were the ones targeted by the EU action. If MOFCOM (or Nuctech) were acting to retaliate against Smiths' use of antidumping to restrict large cargo scanning systems or to help it gain a foothold in this market, it would not exclude these items in the dumping investigation.

From an economics perspective, then, we therefore agree with the Panel's ruling that where the product composition in the bundle of dumped imports and like domestic products differ, adjustments need to be made by an investigating authority to ensure proper comparability. But what about the legal question of whether the vague language of Article 3.2, especially in contrast to Article 2.4.2, is a signal of the treaty drafters' intention to afford authorities with broad discretion to perform price comparisons as they see fit? Is the Panel's ruling too activist in the sense that it chose to create a requirement not found in the text of the law itself?

While we are mindful that WTO panelists ought to interpret treaty provisions in strict adherence to the Vienna Convention on the Law of Treaties, in this particular instance, we think that the Panel did not overstep its authority. Instead, the textual hook can be found in Article 3.1 of the ADA. That provision suggests that investigating authorities, when making an injury determination (which includes the price undercutting and price suppression analyses at issue in this case), must do so on the basis of an ‘objective’ examination. To not take into account the product composition when large price differences exist across the set of goods that fall within the ‘like’ products basket, in our opinion, falls outside the bounds of what, in our opinion, is considered to be objective. Thus, the Panel was right to place some boundaries on the scope of the discretion enjoyed by investigating authorities when undertaking a price comparison of dumped imports against domestic like products.

This then leads to a basic question lurking in this case: Was the approach taken by MOFCOM done simply as a matter of administrative convenience? Or was it part of a concerted strategy to increase the chances of finding injury to further a broader industrial policy strategy? Under the former interpretation, MOFCOM was acting innocently; relying on aggregate data and simple average unit values for domestic and foreign sales allowed for cheap and simple comparisons. Alternatively, MOFCOM could have chosen this methodology deliberately so as to increase the likelihood that injury would be found to a domestic industry.

The Panel Report hints at suspicion that this choice might have been taken with forethought. It notes that Nuctech had provided, on its own initiative, price data on a model-by-model basis, allowing MOFCOM to easily compare prices by model.Footnote ¹²⁸ All MOFCOM had to do was to ask Smiths to provide comparable model-by-model pricing data. But MOFCOM chose not to ask the respondent. To not follow up when the information has already been provided by one party suggests that MOFCOM may well have intentionally chosen to avoid making price comparisons of this form.

Therefore, the Panel declared that MOFCOM's failure to make adjustments, despite being aware of the significant differences in uses, physical characteristics, and prices of high- and low-energy scanners, to ensure proper comparability amounted to a violation of ADA Articles 3.1 and 3.2.Footnote ¹²⁹ If it is the case that MOFCOM's actions were taken with forethought, then this example demonstrates how government authorities will resort to manipulating technical aspects of an antidumping investigation to help push forward an industrial policy. As we suggested earlier, the resort to WTO litigation will then focus on highly technical issues – such as those related to product market definition and price comparison calculations under Article 3 – in an attempt to affect the impact of antidumping measures and strategic industrial policy.

Importantly, the Panel ruling leaves unresolved several important questions regarding price comparisons for an Article 3 injury analysis under the ADA. Future WTO Panels, and eventually the Appellate Body, should assist with providing greater clarity on three sets of questions.

First, the Panel simply finds MOFCOM's behavior to be problematic. It does not clarify what types of adjustments, had they been made, would have sufficed to satisfy the law. Would it have been fine for MOFCOM to simply exclude high-energy scanners? Could it have otherwise used the same exact calculation methodology if it had limited the aggregate data to only low-energy scanners? Or must it have undertaken price comparisons on a model-by-model basis? Did it need to undertake a transaction-by-transaction analysis? Because the Panel only states that adjustments are necessary but does not clarify what these must entail, investigating authorities are left to speculate until the question is clarified in another future case.

Second, the Panel notes that even if high- and low-energy scanners are ‘like’ products, adjustments were warranted in light of the differences in three factors: (1) product characteristics; (2) use; and (3) price. Must there be differences along all three dimensions before an adjustment is necessary, or would differences across a subset of these factors have also sufficed? More specifically, is it critical that the uses of A and B be different? The Panel's decision to emphasize this difference suggests that it may be. But suppose A are luxury watches and B are non-luxury watches. Their use is nearly identical, although their physical characteristics and prices differ. Under this scenario, if the product mixes across dumped imports and domestic ‘like’ products differ, are adjustments also necessary? One might think that the answer is also yes. But the Panel ruling leaves unclear what factors must be considered and how any multifactor test is to work before the requirement to make adjustments to ensure comparability is triggered. Greater guidance for investigating authorities in terms of knowing when they must divide a basket of ‘like’ products and how fine the divisions must be would be helpful.

Third, the Panel ruling leaves unclear what burdens fall on the investigating authority in terms of seeking information to make adjustments. The Panel faults MOFCOM for not seeking model-by-model data from Smiths when this information was provided by Nuctech. The ruling therefore clarifies that where one party has submitted detailed information which allows for adjustments to be made, the investigating authority has a burden to seek comparable information from the other party. But what if this is not the case? Suppose, for example, that neither party had submitted data on a model-by-model basis. After all, it is not altogether clear that Nuctech should have done so proactively, as it did not serve its interests. Would MOFCOM, lacking sufficient data from both parties, then have been free to undertake the calculations as it did? Or would it nevertheless have been required to ask for data to make adjustments because it could already tell, based on HS-8 customs data, that the product mix of Smiths' imported products might be different than that of Nuctech? Again, greater clarity would be helpful.

To summarize, the China–X-Ray Equipment dispute helps to clarify some portion of the law with respect to product comparisons and price comparability in the injury analysis of an antidumping investigation. The focus on this element of the law is not altogether surprising. Manipulations of such calculations, as we have suggested, may serve as a means to bolster support for an industrial policy through antidumping. Furthermore, given the lack of success with challenges of the ‘like’ product definition under ADA Article 2.6, WTO members are choosing to make narrower claims, as the EU did in this case. If we are correct in our assessment, then we should expect more disputes with ADA Article 3 claims to surface in the coming years. The Panel's ruling in China–X-Ray Equipment certainly does not offer a definitive last word on this issue. In spite of the Panel's efforts, much of the law on this issue remains in need of further judicial clarification.